Understanding recent changes in SOFR-based loan index rates

Three months into 2022, the transition from LIBOR to SOFR as a base rate for new originations seems close to complete. We’re no longer seeing new loans indexed to LIBOR, and the only open question seems to be if the lending market will coalesce around a single type of SOFR or if we’ll continue to observe different lenders using different “types” of SOFR (though there are still a lot of questions on how lenders plan to transition their legacy LIBOR books).

SOFR loan originations are using one of three different "versions" of SOFR. Freddie Mac and Fannie Mae floaters are using New York Fed 30-Day SOFR, which can be thought of as an average of the daily SOFR rate over the month prior to a loan’s current interest period. Most non-Agency lenders are now using 1-month Term SOFR, which is derived by looking at where SOFR futures contracts trade relative to spot SOFR, reflecting what the market "expects" today that daily SOFR will average over the coming month. An increasing minority of non-bank lenders are also using daily simple SOFR, which essentially averages daily SOFR rates over the interest period (with the downside being that the month’s interest rate and expense aren’t then known until the end of the month).

Until recently, the choice between these different rates in a loan hasn’t appeared to matter much. As the table below shows, these three rates have largely been the same, suggesting that using one versus another was irrelevant — regardless of which one was used in a loan, the interest expense would be the same.

In recent weeks, this appears to have changed. In anticipation of the Fed raising interest rates, Term SOFR moved up even as spot SOFR and New York Fed 30-Day SOFR remained relatively unchanged. We’ve heard from many borrowers with Term SOFR loans that they are concerned that their rates will be permanently higher on the order of 25 basis points because of the version of SOFR that they’re using.

The explanation for this divergence lies in how these different SOFR rates are determined. Spot SOFR derives from where overnight repo transactions clear, which themselves tend to be bounded by the current Fed Funds target rate. When the Fed Funds target rate was at 0–25 bps, spot SOFR has stayed within this range (closer to the lower part of it), with New York Fed 30-Day SOFR tracking this on a lagging basis. 1-month Term SOFR, however, is a forward-looking rate. It looks at where SOFR futures contracts trade relative to current SOFR and, in essence, captures what the market expects spot SOFR to average over the coming month. Put more succinctly, Term SOFR reflects expected market conditions over the coming month, daily SOFR reflects current market conditions, and New York Fed 30-Day SOFR reflects past market conditions.

The steady increase in Term SOFR shown above started right when the Fed meeting (and the presumed rate hike) was one month out (the jump to 10 bps on February 10 was due to short-lived expectation of an inter-meeting hike when higher than expected inflation data came out). This forward-looking rate behaved exactly as it should — it started to rise in anticipation of spot SOFR increasing when the Fed raises rates. With the passing of each day, the forward-looking, one-month period reflected in Term SOFR included more and more days after the expected hike, so it continued to rise. Spot SOFR (which reflects current market conditions) and New York Fed 30-Day SOFR (which reflects market conditions on a trailing 30-day basis) didn’t move because the Fed hadn’t hiked rates yet.

Through this lens, the increase in Term SOFR leading up to the March Fed meeting makes sense, as does what happened afterwards when the Fed increased the Fed Funds target rate from 0-25 bps to 25-50 bps. On the day immediately after the rate hike, spot SOFR jumped from 5 to 30 basis points, where it has stayed since. New York Fed 30-Day SOFR has started to slowly increase, as you would expect from a number that is an average of spot SOFR over the previous 30 days. We would expect it to hit 30 basis points (the current level of the other two SOFR indices) by mid-April, though by that time 1-month Term SOFR may start rising again in anticipation of another rate hike from the Fed at their May 4 meeting. This will be interesting to watch as the market is pricing for a 50 bps hike at that meeting (as of this writing, the market was pricing in a 68% chance of another 50 bps hike in that meeting and a 32% chance of a 25 bps hike); if this continues to be priced in as we get closer to May 4, we could see Term SOFR increase even more dramatically relative to spot SOFR than we saw leading up to the March 15 rate hike.

This explanation should help a borrower get comfortable that even if Term SOFR is around 25 basis points higher than spot SOFR or New York Fed 30-Day SOFR at times in anticipation of rate increases, that gap won’t likely persist over the life of the loan. The forward curves below, which show market implied future rates for Term SOFR and New York Fed 30-Day SOFR, reflect this fact — they are quite similar. The difference between different SOFR base rates is likely not a “blip” as the current forward curve shows. The market is pricing in the probability of several additional rate hikes; seven more in 2022, and two more in 2023. If those rate hikes occur (which is a big “if”), we’d expect to see the same dynamic play out with the different versions of SOFR. 1-month Term SOFR would start to steadily rise about a month out from any expected hike; spot SOFR would remain flat but increase immediately after the hike by approximately the amount of the hike. New York Fed 30-Day SOFR would steadily rise over the following month. This would play out each time the market priced in a hike for a Fed meeting and the hike actually occurred. Situations where a hike was not anticipated but occurred anyway, or where it was anticipated but did not occur, would result in different dynamics.

These forward curves may be used to calculate swap rates for Term SOFR and New York Fed 30-Day SOFR, which reflect what the market “expects” each rate to average over the coming years. While actual rate setting will likely differ from what’s implied by these curves and swap rates, they do let the borrower know what the market expects the difference between different types of SOFR to be over an extended period of time. The chart below compares swap rates for 1-month Term SOFR with New York Fed 30-Day SOFR. This is a good comparison as many of our clients are multifamily investors who can choose between Agency floating-rate debt (which uses New York Fed 30-Day SOFR) and transitional loans from non-bank lenders (which largely use Term SOFR now). There's a lot more that goes into this choice than just the index rate (not the least of which is proceeds!), but it's a comparison we're often asked to help explain.

It's clear from these numbers that the market is pricing in 1-month Term SOFR to average higher than New York Fed 30-Day SOFR, but the spread between the two clearly declines as the time horizon gets longer. Over a one-year period, the difference is expected to be 17.7 basis points, declining to about 9 bps over a two-year period and about 5 bps over a three-year period. An investor that closes a loan indexed to Term SOFR might reasonably underwrite an additional 5 bps of interest expense on a running basis over three years compared to an otherwise identical loan indexed to New York Fed 30-Day SOFR. This difference is entirely dependent on the rate environment — the numbers above indicate that the market is pricing in quite a few rate hikes (~8 hikes of 25 basis points each in the next two years) and that Term SOFR captures these hikes on a leading basis while New York Fed 30-Day SOFR reflects them on a lagging basis. If the market was pricing in rate cuts, the opposite relationship would be observed — Term SOFR would be lower than New York Fed 30-Day SOFR on an expected forward-looking basis.

As general rules of thumb:

• Term SOFR and daily simple SOFR will be quite close in any given interest period so long as market anticipated increases/decreases in rates occur as predicted.
  

• These rates won’t match if an expected increase/decrease in rates doesn’t occur as anticipated.

• Term SOFR and New York Fed 30-Day SOFR will differ in periods when the Fed hikes/cuts rates. Specifically, Term SOFR will generally anticipate rate hikes/cuts by the Fed by an interest period while New York Fed 30-Day SOFR will reflect those hikes/cuts on a lagging basis; this could result in interest expense on a Term SOFR loan being at least marginally higher in a rising rate environment relative to a loan with New York Fed 30-Day SOFR; conversely in a declining rate environment, a loan indexed to New York Fed 30-Day SOFR would be expected to have greater interest expense than a similar loan indexed to Term SOFR

We hope this explanation provides a better understanding of this nuance of SOFR indices. As always, please reach out to us with any questions. If you'd like us to take a look at a particular loan agreement, or want to better understand how the cost of a cap might differ depending on the SOFR index being used, we’d be happy to discuss in detail with you.